Real World Data – Realistic Population -- Evidence for Unrecognized Pharmacokinetics of Renally Eliminated Drugs
Nick Holford (1), Conor J O’Hanlon (1), Karel Allegaert (2), Brian Anderson (3), Amilcar Falcão (4), Nicolas Simon (5), Yoke-Lin Lo (6), Alison H Thomson (7), Catherine M Sherwin (8), Evelyne Jacqz-Aigrain (9), Carolina Llanos-Paez (10), Stefanie Hennig (11) , Linas Mockus (12), Carl Kirkpatrick (13)
1. Department of Pharmacology & Clinical Pharmacology, University of Auckland, New Zealand; 2. Department of Development and Regeneration, KU Leuven, Leuven, Belgium & Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Leuven, Belgium & Department of Hospital Pharmacy, Erasmus MC, Rotterdam, The Netherlands; 3. Department of Anaesthesiology, University of Auckland, New Zealand; 4. Faculty of Pharmacy, Coimbra Institute for Biomedical Imaging and Translational Research, University of Coimbra, Portugal; 5. Aix Marseille Univ, Hop Sainte Marguerite, Service de Pharmacologie clinique, Marseille, France; 6. Department of Pharmacy Practice, School of Pharmacy, International Medical University & Malaysia & Department of Pharmacy, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia; 7. Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, United Kingdom; 8. Department of Pediatrics, Wright State University Boonshoft School of Medicine/Dayton Children’s Hospital, Dayton, Ohio, USA; 9. Paediatric Pharmacology and Pharmacogenetics, APHP Hôpital Saint-Louis – University Paris Cité, Paris France; 10. Department of Pharmacy, Uppsala University, Uppsala, Sweden; 11. Certara, Inc., Princeton, New Jersey, USA & School of Clinical Sciences, Faculty of Health, Queensland University of Technology, Brisbane, Australia; 12. Chemical Engineering Department, Purdue University, Indiana, USA; 13. Monash Institute of Pharmaceutical Sciences, Faculty of Pharmacy and Pharmaceutical Sciences, Monash University, Melbourne, Australia
Objectives:
- Create a population dataset from real-world data sources describing humans from premature neonates to elderly adults.
- Identify obstacles combining data from multiple sources to obtain a coherent and consistent population dataset.
- Use real-world data to describe the pharmacokinetics of renally eliminated drugs.
- Describe how size, age, and renal function influence renal clearance components and volumes of distribution.
Methods: Real-world data was obtained through routine clinical blood sampling used to describe gentamicin, amikacin, and vancomycin exposure from 18 hospital and university-based sources worldwide. Patient demographics, clinical data, drug dose size, time and routes of drug administration, sampling time, and measured drug serum concentrations were obtained at each source from medical record databases using source-specific data extraction procedures. The primary challenge was combining data from each source. This was overcome by establishing an event format based system which captured many kinds of data items with their original units. This facilitated conversion to a consistent NONMEM data format. Criteria were established to exclude implausible total body mass values (based on associated age, sex, and height) and implausible renal function values. Missing height and renal function data were imputed.
A mixed effects joint model of the pharmacokinetics of these 3 drugs was developed from the real-world population data using NONMEM 7.5.1. Details of the calculation of fat-free mass, creatinine production rate, creatinine clearance, estimated GFR, normal GFR, and renal function (RF) are outlined in reference (1). The GFR clearance component was assumed to approach an asymptotic value determined by normal GFR, while the non-GFR clearance component was not constrained by normal GFR. Normal fat mass (2) was estimated for both clearance components and for distribution parameters to account for differences in body composition. Individual differences in drug clearance were described by body size, composition, maturation, and renal function (defined as the ratio of estimated GFR to normal GFR).
Results: The exclusion criteria removed 12.6% of the original pooled real-world data because of implausible values (10.9% due to size and 1.7% due to renal function). The final pooled dataset comprised 27,341 drug concentrations in 9,951 patients. This extensive real-world dataset provided a wide dispersion of individual patient covariates. Gentamicin, amikacin and vancomycin are thought to be predominantly excreted by the kidneys.
A physiological approach was employed to describe total clearance from premature neonates to elderly adults (3). This study (3) demonstrated that GFR is a predictor of drug elimination clearance with two distinct components: GFR clearance associated with normal GFR (CLGFR) and clearance not associated with GFR (CLNGFR). This distinction between CLGFR and CLNGFR clearance components is analogous to using saturable and non-saturable binding as a function of unbound concentration to describe total plasma protein binding. The maximum binding capacity is analogous to normal GFR and non-specific binding is analogous to CLNGFR.
All three drugs had CLGFR estimated as a drug-specific percentage of normal GFR (gentamicin 39%, amikacin 88%, vancomycin 55% in as standard adult with RF=1). Drug-specific maturation of size and composition-scaled volume of distribution was observed. There was an initial increase at birth relative to an adult for both central volume (58% gentamicin, 35% amikacin, 4.7% vancomycin) and peripheral volume (1.5% gentamicin, 25% amikacin, 2.5% vancomycin). The maturation of volume fell exponentially to just 5% above adult values by 3 years (central volume) and 3.2 years (peripheral volume).
Conclusions: Pooling data from multiple real world data sources was essential to achieve a wide dispersion of size, age, and renal function. These real world data have been instrumental in distinguishing previously unrecognized clearance components for drugs with extensive renal elimination and in describing the maturation of central and peripheral volumes of distribution starting at birth. Recognition of these pharmacokinetic changes across the human age span has immediate applications for models of renally eliminated drugs using target concentration intervention (4) and for implementation of rational dosing during clinical drug development.
References:
- O'Hanlon CJ, Holford N, Sumpter A, Al-Sallami HS. Consistent Methods for Fat Free Mass, Creatinine Clearance and Glomerular Filtration Rate to describe Renal Function from Neonates to Adults. CPT: pharmacometrics & systems pharmacology. 2023;12:401-12.
- Holford NHG, Anderson BJ. Allometric size: The scientific theory and extension to normal fat mass. Eur J Pharm Sci. 2017;109(Supplement):S59-S64.
- Holford N, O'Hanlon CJ, Allegaert K, Anderson B, Falcão A, Simon N, et al. A physiological approach to renal clearance: From premature neonates to adults. Br J Clin Pharmacol. 2024;90(4):1066-80.
- Holford NH, Ma G, Metz D. TDM is dead. Long live TCI! Br J Clin Pharmacol. 2022;88(4):1406-13.
